30 Oct Personalization: Learning, performance – and the Pingu effect
Online personalization – content served to the end user based on what it deems suitable for you – isn’t always perfect. But applied to learning it offers incredible opportunities for funnelling precise information to the right learners. Learning Designer Lauren Keith takes us into the future and back to basics…
It can be quite a wakeup call to see your ‘recommended products’ on Amazon, as based on your purchase and browsing history. Looking at mine, they have me profiled as a compulsive horror fan, who’s always losing phone chargers, has an unfounded obsession with shampoo and watches a lot of Pingu. They’re 75% right.
The Pingu DVD I bought last month was a one off gift for my niece. But even Amazon’s big data engine couldn’t know that.
Personalization for consumers
Most of us have experienced personalized content as a consumer. On YouTube, you’re recommended songs by similar artists to those you’ve recently listened to previously. On supermarket websites, you may be reminded of those staple food items you tend to buy every week but could easily forget. And of course on Amazon, you’re pointed to new products that appeal to your presumed needs and interests. So based on my backlog of horror films, I’m unlikely to be recommended a Jennifer Aniston rom-com or the latest period drama.
Personalization can use information from all sorts of things, from the device you’re using (be it phone, tablet or PC), the location you’re accessing it from (whether in town or at home, for example), your user history (past purchases, posts), session behaviour (page view, navigation clicks) and so on.
The data absorbed can be used to do something simple – like adapting your screen layout to be mobile-ready, or something much more sophisticated. For example, Google has recently introduced artificial intelligence and machine learning to its search engine. Its system, RankBrain, uses its ability to understand human language and continuously learn to interpret what information Google search users are looking for. It then generates results tailored to that user’s particular request. In other words, it knows what you’re looking for better than you do!
The benefits of personalization? As the internet and entertainment industry continue to be overwhelmed with more information than ever before, providing goal-orientated content is becoming a more and more valuable service.
Sifting through infinite content can be time-consuming and tiresome, and when 74% of online consumers get frustrated with websites when content appears that has nothing to do with their interests, you can see why marketers employ personalization.
The downfalls? The assumptions made by data engines to bring you personalized content may not always be completely accurate. For a start, the Pingu merchandise on my ‘My Amazon’ page is taking up valuable screen space that could otherwise be used to market something more suitable for me, just because of that one anomalous purchase. Secondly, what if I might like Jennifer Aniston films? I could be missing out on some great new film titles that cater to the romantic, happy-go-lucky side of my personality.
When we apply this to learning it gets really interesting. What is personalized learning and how could it enrich our learning experiences?
Personalized learning in a nutshell
In terms of learning, personalization works in a similar way to the consumer-marketing dialogue. The content the website software deems unhelpful for that learner is filtered out in favour of what it thinks is best suited to them. At the most basic level, this could be done by job role. The website or LMS may already have information about each learner’s position which it uses to hide and/or display information. Alternatively, it can ask them for this data directly.
Taking ‘personal’ further, the content could be filtered by the learner’s experience, existing skills, preference for learning or competencies. We will explore this later.
Why personalize learning?
Being able to target learners with specific material instead of having a one size fits all deliverable has a number of benefits. Firstly, consider it from a practical perspective. Adult learning theorists frequently tell us adults place huge importance on the relevance of information and how immediate its application will be. Having content specifically curated for us, or geared towards a specific goal is going to dramatically increase our chances of accessing and digesting that content – and indeed, having the time to do so.
Secondly, being able to cater to a particular learner’s ability can be instrumental for motivating them. Content that’s too advanced for us can stifle our confidence and with it, our drive for learning. On the other hand, content that’s too easy can cause frustration or lead us to believe we’re wasting our time. Having content set to our particular level avoids both these obstacles.
Finally, personalized content can be more empathetic and encouraging for the learner. The same information, for instance, could be re-produced to take a different perspective or emphasis depending on the target audience. It means getting on the learner’s level and striking the right emotional chords before dispensing the knowledge, prepping the learner to be more receptive and to process the vital information more thoroughly.
What about the Pingu effect?
That’s all very well but what if personalized learning isn’t accurate? What if, like the Pingu merchandise, I’m being pushed content I don’t need? Or, like the Jennifer Aniston movies, I’m missing out on something, just because of the pigeon-hole – or penguin-hole – the website has put me in?
Let’s look at three ways that personalized learning can account for this.
1. Letting the learner pick their own learning pathways
Curators and administrators may want to set out learning journeys for learners based on shared goals or training needs. But these don’t have to restrict the information available to the learner, or be compulsory. Brightwave’s next gen LMS, tessello, gives the option for the learner and administrator to co-create structured learning pathways by collating appropriate learning materials that may be helpful to achieve a particular goal, and publishes them as a pathway under a particular heading.
Any learner searching this pathway can use it for as long as it suits their needs – but that’s not to stop them searching and accessing other learning pathways – or other more general content, and using these to change the pathway’s course. It’s about providing direction – optional pathways as opposed to locked doors.
2. Using opt-in data
When I bought the Pingu DVD, I would have liked an option to say ‘I do not want Amazon to remember this purchase’ and so avoid being plagued with Pingu products every time I log in. By the same principle, when it comes to learning, giving the learner an element of control over the data that’s input to the website to generate their content can be beneficial.
Websites can use opt in information. That is, information is asks the learner to provide, rather than simply taking it from their search history and so on. This could be in the form of a diagnostic questionnaire or some other form of pre-qualification exercise.
For instance, learners can be presented with a pre-learning survey for which the results determine the learning pathways served. This can work like an informal test, asking knowledge based questions, or like an attitudinal one, asking the learner what they perceive their own abilities to be.
Either way, the learner has a degree of control over the data they provide and understands why they’re doing so. Either way, it’s better for engagement and making the learner receptive to learning content.
3. Choosing your rules carefully
There may be times where personalization should require no effort on the learner’s part and where they don’t even need to think about how the content is filtered and served. This requires smart diagnostic tools equipped to consider multiple complex variables – likes and interests, responsibilities, time pressures, preferences for different media, as well as the basics like job role competencies, and capability frameworks.
This is something that will only improve in the future. We are moving towards a point where technology can instantly distinguish between anomalous and significant behaviour, understanding the multi-faceted and often contradictory components of ourselves and our situations – Pingu and all – and serve us up exactly the right drop from the oceans of online learning content – before we’ve even realised we need it.